论文部分内容阅读
电能、燃气、热能等多能源融合使能源在生产—调配—使用过程中极易发生多能扰动,传统的离线扰动识别方法很难满足能源互联网实时性的要求,采用数据流处理技术可对多能扰动信号进行在线识别。从能源互联网的能量—信息交互入手,对能源互联网多能扰动问题进行分析,为不同扰动信号建立数学模型。采用小波变换对扰动信号进行分解,并基于滑动窗口构建扰动信号的数据流处理模型。该模型首先构建滑动窗口概要数据结构,其次改进小波树更新算法以实现概要结构快速更新,优化扰动信号特征提取,最后采用决策树算法对信号特征进行分类。构建的数据流处理模型被应用到电能质量扰动和燃气质量扰动的识别中,验证该数据流处理模型的有效性。
Multi-energy fusion such as electricity, gas and heat energy makes the energy prone to multiply perturbation in the process of production-deployment-use. The traditional offline disturbance identification method is difficult to meet the real-time requirements of energy Internet. Can disturb the signal for online identification. Starting with the energy-information interaction of energy Internet, this paper analyzes the problem of energy-perturbation of energy internet and establishes mathematical models for different disturbance signals. The wavelet transform is used to decompose the disturbance signal and construct the data flow processing model of the disturbance signal based on the sliding window. The model first constructs a sliding window summary data structure, then improves the wavelet tree updating algorithm to realize the quick update of the summary structure, and optimizes the feature extraction of the disturbance signal. At last, it uses the decision tree algorithm to classify the signal features. The constructed data flow processing model is applied to the identification of power quality disturbance and gas quality disturbance to verify the validity of the data flow processing model.